Back home
Retrieval

AI knowledge assistants, answering from your documents.

The answer, and the page it came from.

Your organisation already knows the answer. It is in a policy nobody can find, a contract in a shared drive, a ticket somebody closed two years ago. We build assistants that search that material and answer from it, with a citation on every response so the reader can check the source rather than trust the tone. These assistants answer and cite. They do not take actions in your systems: that is agent development, on its own page.

Where a knowledge assistant earns its place

The pattern is always the same, the value is not. These are the places where the search that already exists is failing people daily.

01

Internal help desk and policy assistant

The questions your operations, HR, and finance teams answer over and over: expenses, leave, procurement rules, approval thresholds. Answered from the current policy, with the clause attached, so nobody has to remember which version is live.

  • HR, expense, and procurement policy
  • Internal IT and onboarding questions
  • Answers linked to the governing clause
  • In Slack, Teams, or your intranet
02

Product and technical documentation

Documentation people can ask rather than browse. Useful where the answer sits across a manual, a release note, and a known-issue list, and where the reader does not know which of the three they need.

  • Product manuals and specifications
  • API and developer documentation
  • Release notes and change history
  • Troubleshooting and known issues
03

Contract, tender, and compliance lookup

Long documents where the answer is one clause and finding it is an afternoon. The assistant retrieves the passage, quotes it, and links the source, so the judgment stays with the person and only the searching is automated.

  • Clause and obligation lookup
  • Tender and bid requirement checks
  • Regulatory and standards reference
  • Comparison across document versions
04

Assistants over ticket and case history

Support and field teams solving something that has been solved before. Retrieval over closed tickets and case notes surfaces the fix and the case it came from, which shortens handling time and stops the same escalation twice.

  • Similar cases and their resolutions
  • Suggested replies drafted for review
  • Escalation history and context
  • Inside your existing helpdesk
05

Customer-facing answer engines

The same system pointed outward, with a tighter boundary: only published material is in scope, answers carry links, and anything outside the material goes to a person instead of being improvised.

  • Support answers from published help content
  • Pre-sales product questions
  • Handoff to a human with context kept
  • Gaps in the content reported back
06

Field and frontline assistants

For people who are not at a desk. Short answers on a phone, from service manuals and site records, built to work with a scanned part number and a poor connection.

  • Service and maintenance procedures
  • Parts and equipment lookup
  • Site and asset history
  • Mobile, with offline caching
How we work

How the work runs

  1. Tell us what you need

    A short written brief is enough to start: what you are building, the stack it has to live in, and the problem behind it. Someone technical reads it, not a sales rep.

    passed when we reply with questions and a proposed shape

  2. Scope and quote

    We agree what the first release contains and what it costs before anything is built. Anything we cannot estimate honestly gets scoped on its own rather than padded into the total.

    passed when the scope and the quote are agreed

  3. Build in the open

    Working software arrives in increments, in your repository, with our commits visible from the first week. Progress is something you can open and run, not a status report.

    passed when a real user is doing a real job in it

  4. Iterate, then hand over

    What the first release teaches sets the priority for the next one. When you want it in house, documentation and a handover period are part of the engagement rather than an extra.

    passed when your team can run it without us

How pricing works

when the scope is clear

Fixed price per milestone

You approve each milestone and its price before it starts, so the cost is known before the work is rather than after. Milestones are sized to be worth reviewing on their own.

when the work is ongoing

Monthly rate for a dedicated team

A set team at a set monthly cost, with notice rather than a minimum term. Priorities move between sprints without renegotiating anything.

when the shape is still open

Scoping first, quote after

We work out what it should be with you before quoting it. There is no charge for that conversation and no obligation to go ahead.

Frequently asked questions

What makes the answers trustworthy

Every build ships end to end, from first scope to live deployment and ongoing support.

Ingestion from the systems the documents already live in
Hybrid search, so exact terms and loose phrasing both work
A citation on every answer, linked to the source
Your permissions carried through retrieval, per user
Re-indexing on a schedule, so answers follow the current version
An honest "I do not know" instead of a confident guess

Send Inquiry

Regarding: AI knowledge assistants
What are you looking for? optional, pick as many as apply

By submitting this form, you agree to our Privacy Policy and Terms of Service.